Phishing is a very serious challenge in web-based security these days. Phishing websites are forge web pages that are designed by noxious individuals to emulate pages of genuine sites. Phishers regularly make website pages that looks outwardly like the authentic and original site pages used in deceiving their victims. An uninformed internet customer is likely to be effortlessly tricked unknowing that this is a scam. Phishing webpages victims may uncover their financial details, secret key, and other sensitive data to the proprietor of phishing webpage. This research work is exceptionally significant in that it aimed at developing a model that will give security to users from phishers fraudulent tricks, and enable the users to distinguish the authenticity of websites. The benefit of this research work is to enable government institutes and financial organization to render variety of secured financial services to their various customers. The proposed model used association rule mining algorithm called Predictive Apriori to address phishing website issues that has been a global threat.  The algorithm is proficient in analyzing and extracting phishing URL features and is able to perform URL feature classification to identify phishing websites. It then flags the website as either legitimate or malicious. Therefore the user will be able to open only legitimate links. Systems Analysis and Design Method and Prototype methodologies were applied during this research. Our results showed that the algorithm used in this research work generates more accurate predictions than that of the existing system